New overview adds more evidence for safety of third generation oral contraceptives

July 13, 2011

A new overview of studies of the Pill and the risk of heart attack, published today (Friday 30 August) in Europe's leading reproductive medicine journal, Human Reproduction,* provides a reassuring picture of the safety of the third generation Pill.

Reprinted with permission of ESHRE

A new overview of studies of the Pill and the risk of heart attack, published today (Friday 30 August) in Europe's leading reproductive medicine journal, Human Reproduction,* provides a reassuring picture of the safety of the third generation Pill.

A joint Canadian-British** team aggregated seven recent epidemiological studies that investigated the risk of myocardial infarctions (MIs) in users of second and third generation oral contraceptives (OCs)***.  They used the well-established method of Petitti**** to derive pooled estimates.

They concluded that the recent studies consistently demonstrated the increasing safety of both types of Pill - and the third generation Pills in particular - as far as heart attack risks are concerned.

Lead investigator Professor Walter Spitzer of McGill University in Montreal, explained:  "Recent controversies about OCs, in particular third generation OCs, have dwelt upon blood clots in the veins.  Less attention has been paid to the safety record of newer OCs in respect to the arteries, and all except one of the recent studies lacked statistical power as stand-alone projects.  So, we aggregated the findings of seven studies involving nearly 6,500 women since 1996.

"The data corroborate that all oral contraceptives currently on the market formally studied since 1996 as risk factors for MI are safe when used according to their regulatory labels.  The absolute rates of occurrence of MI among users in reports since 1995 are low - unequivocally lower than those reported earlier between 1966 and 1995."

The researchers concluded that a worst interpretation of their overview still favours third generation OCs in the risk of a heart attack. The worst relative risk (RR) for third generation OC use was found to be 0.62 in comparison to second generation OC. [An RR of 1.0 indicates no effect, greater than 1.0 harm, and an RR less than 1.0 indicates benefit]. The authors showed that the range of relative risks of third generation OCs versus second generation OCs extends from 0.43 to 0.62, all statistically significant. They believe that 0.44 is the most rigorous estimate because the synthesis incorporates only the studies least affected by any possible bias or design problems. Four different permutations and combinations of the seven studies were done in order to minimize any effect of bias.

Because MI is so rare in women of reproductive years, the absolute rate of MI was important to focus on. It has become extremely low, whether or not women are on any Pill and whether or not they took second or third generation OCs.  Because only three of the seven post 1996 studies reported absolute MI rates, a formal aggregation was not appropriate. However, a cautious and conservative estimate suggested a rate of only 1.3 (range 0.6 -1.8) per 100,000 women years in combined second and third generation Pill users and non-users. By comparison, in pre 1996 studies it was 13 per 100,000 women years in Pill users compared with 1.5 in non-Pill users. (The research team had also evaluated 22 studies published between 1965 and 1995).

Said Professor Spitzer: "With respect to arterial adverse events in relation to third generation OCs, the recent literature offer grounds for reassurance, specially in regard to acute MI. That is true whether the interpretation is either benefit or no difference between third and second generations or between the third generation of OCs and 'no use'.

"Most importantly, in the early studies published before 1996, a median RR just above 2.0 for any OC then on the market was observed. In the more recent studies we observed pooled risks of 1.11 for the four most rigorous studies comparing third generation OCs with no use. For second generation OCs versus no use the pooled RR is 2.54. Attainment of no difference between the risks of exposure to third generation OCs and non-use of any OC product in respect to myocardial infarction has allowed the laboratory investigators to realize their goals of developing newer and safer preparations. This is the first epidemiological report of such equivalence.

"Subsets of older women - those with a family history of MI or an unfavorable profile of cardiovascular risk factors --- should be candidates for third generation OCs. Smoking, whether or not it was adjusted for OC use, is a very strong risk factor for MI. If a woman cannot or will not cease smoking and insists on oral contraception, the data favour third generation OCs. Women with elevated blood pressure should be prescribed OCs only with great caution and after it has been brought down.  That's true for both second and third generation Pills."

He also stressed that the findings suggested that it would be imprudent to curtail the armamentarium of any approved OC for first-line use.

"Our conclusions should not be interpreted as recommendations against second generation Pills or as a strong recommendation in favor of third generation Pills. With informed consent by the counseled woman, the choice of an approved OC should always be that of the counseling clinician, based primarily on clinical judgment, one patient at a time."

* Myocardial infarction and third generation oral contraceptives: aggregation of recent studies. Human Reproduction.  Vol 17.  No 8.  pp 2307-2314.
** Walter O. Spitzer, Emeritus Professor of Epidemiology, McGill University, Montreal, Canada;
Ms Janet M Faith, Department of Epidemiology and Biostatistics, McGill University, Montreal, Canada; Kenneth D MacRae, Professor of Medical Statistics, University of Surrey, Guildford, Surrey, UK.
NB: Prof. MacRae died before publication of this paper.
*** Oral Contraceptives are defined as second or third generation according to the type of progestogen they contain. Second generation OCs: norgestrel, levonorgestrel. Third generation: desogestrel, gestodene [orgestimate is under debate because of its close relationship with levonorgestrel]
**** Petitti DB et al (2000), Meta-analysis, decision analysis, and cost-effectiveness analysis: methods for quantitative synthesis in medicine. Monographs in Epidemiology and Biostatistics. Oxford University Press, New York. Vol 2.

2 Human Reproduction is a monthly journal of the European Society of Human Reproduction and Embryology (ESHRE). Please acknowledge Human Reproduction as a source.
ESHRE's website is:
3 Printed texts available on request from Dr Helen Beard, Managing Editor.
  Tel: +44 (0) 1954 212404 or email: (media inquiries only):
Margaret Willson Tel: +44 (0)1536 772181   Fax: +44 (0)1536 772191   Mobile: +44 (0)7973 853347
Home tel: +44 (0)1536 770851   Email:
PDF of paper also available on request as well as from website.